Communications
Table 1: Results of the enantioselective addition of cyanoacetates 3 to acetylenic esters 4 promoted by
by 6a or 6b, respectively
(Scheme 3). In these adducts, the
orientation of the acceptor toward
the amine complex bearing cyanoa-
cetate anion, 6a or 6b, has a critical
importance for determining the
enantioselectivity. Noticeably, the
enantioface of the N-bound cyano-
acetate anion used for the coordi-
nation of substrate is opposite for
both adducts: Si face for 7a and
chiral catalysts 1 or 2.[a]
Entry Cyanoacetate Acetylenic ester Cat. Solvent
Yield [%][b] Z/E
ee [%] (Z/E)[c]
1
2
3
4
5
6
7
8
3a
3a
3a
3a
3a
3a
3a
3a
3a
3a
3b
3c
3d
3a
3a
3a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4b
4b
4b
1a
1b
1c
1d
1e
1b
1b
2a
2b
2c
1b
1b
1b
1b
1b
1b
toluene
toluene
toluene
toluene
toluene
CH2Cl2
97
97
98
97
97
98
20/1
20/1
23/1
14/1
6/1
86/22
91/52
83/25
79/13
43/04
82/81
72/72
84/10
90/23
81/10
89/n.d.
92/27
95/05
2/1
2/1
Et(Me)2COH 99
toluene
toluene
toluene
toluene
toluene
toluene
toluene
CH2Cl2
CH3CN
96
96
95
98
99
99
99
99
99
20/1
20/1
23/1
20/1
17/1
20/1
1/22 n.d./79
1/2
1/2
9
À
Re face for 7b. Hence, the C C
10
11
12
13
14
15
16
bond formation would yield inter-
mediates 8a and 8b—with opposite
configuration at the benzylic carbon
atoms—which are precursors of the
products (S)-5a and (R)-5a, respec-
tively.
65/79
0/0
[a] Unless otherwise noted, the reactions were carried out by the slow addition of 4 with a syringe pump
to a solution of 3 and the chiral amido catalyst in 5 mL of solvent for 20 minutes and subsequent stirring
for 24 hours. The molar ratio of cyanoacetate/acetylenic ester/catalyst=100:100:1. [b] Yield of isolated
product after column chromatography on silica gel. [c] Determined by HPLC on a chiral stationary phase
using a Chiralpak AD-H column (4.6 mmꢀ250 mm). n.d.=not determined.
For the computational study, we
have used the Ru–Tsdpen complex
1b because it provides slightly
better enantioselectivities than 2a.
We have optimized the structures of
results in terms of enantioselectivity and the Z/E selectivity
(entries 3–5). Choice of the solvent has a significant influence
on the configuration of the products. Nonpolar solvents like
toluene or dichloromethane gave high enantioselectivity,
while polar solvents lead to a dramatic decrease of the
ee values and the Z/E selectivities (entries 2, 6, and 7). The a-
phenyl-a-cyanoacetates (3b–d), which have substituted aro-
matic rings, reacted with 4a in toluene catalyzed by the chiral
Ru catalyst (1b) at 08C for 24 hours and gave the chiral
adduct with 89–95% ee in excellent yields regardless of the
electronic effect of the substituents (entries 11–13). The
reaction of cyanoacetate 3a and the acetylenic monoester
4b with the chiral Ru catalyst 1b gave the corresponding
adduct with excellent E selectivity albeit with modest ee val-
ues (entries 14–16). Thus, our approach provides sufficiently
high ee values with much higher Z/E ratios than those of the
reported reaction promoted by organocatalysts.[9]
1
Scheme 2. NOE interactions observed in the phase-sensitive 2D H–1H
NOESY spectrum (400 MHz, CD2Cl2, 223 K) of the sample prepared by
the addition of 1.5 equivalents of cyanoacetate 3a to the amido Ru
complex 1b. See the Supporting Information for more details. Ts=4-
toluenesulfonyl.
the adducts 7a and 7b and found two viable transition states
(TSS and TSR, respectively) for the C C bond-formation
À
To gain further insight into the reaction mechanisms, we
elucidated the structure of the intermediate, N-bound Ru
catalyst–substrate complexes 6a and 6b in solution
(Scheme 2). The reaction of complex 1b[10] with a-cyanoace-
tate 3a yielded the equilibrium mixture of 6a and 6b in a
reversible and stereoselective manner—as was observed
previously with chiral Ir complexes.[8] From the 13C DEPT
and HMBC spectra measured at À558C, we have located the
positions of nonprotonated carbon atoms of 6a,b in the
13C NMR spectrum. The deprotonated anionic carbon atoms
resonate at d = 58.7 ppm and d = 56.5 ppm in 6a and 6b,
respectively. Whereas the nitrile carbon atom is notably
shifted downfield compared to free substrate 3a (d =
136.7 ppm in 6a versus d = 117.3 ppm in 3a). These data
confirm N-binding of the substrate in 6a,b.[8]
step that would each proceed to the intermediates 8a,b
(Scheme 3, Figure 1). Formation of either 7a or 7b is
exothermic; they have similar structures appropriate for the
concerted activation. However, 7b is 9.4 kcalmolÀ1 less stable
than 7a. This difference in the stability originates from the
steric interactions of the loose carbomethoxy group with the
substituents on the anionic carbon atom. Similarly, owing to
the smaller steric interactions of the tert-butoxycarbonyl
À
substituent, the transition state of the C C bond-formation
step resulting in the S product is 9.7 kcalmolÀ1 lower in energy
compared to the unfavorable transition state that leads to the
R product. Thus, we have found a structural reason for the
strong bias towards the enantioselective formation of the
S product that is in accord with the experimentally deter-
mined sense of enantioselection. The experimental ee values
are, however, lower than might be expected in view of such a
substantial energy difference. This outcome might be
In computational work we have located adducts 7a and 7b
in which the substrate 4a is activated in a concerted manner
8158
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 8157 –8160